Measurement of jobless total to be improved

No economic statistics are more sensitive or less trusted than the unemployment figures. The Office for National Statistics will this week announce big improvements to the measurement of the jobs market. Diane Coyle, Economics Editor, reports.

In its dying days the Conservative government enjoyed one bright spot each month with the announcement of a big fall in headline unemployment. The trouble was that nobody believed the figures. Even on its own admission, there had been 10 changes to benefit rules since 1979 which had tended to reduce the number of unemployment benefit claimants, and critics counted far more fiddles than that.

On Thursday the Office for National Statistics is due to announce a series of improvements to the monthly figures on the state of the jobs market, to be implemented when the unemployment numbers for September are published next week.

The announcement will not, however, include a switch from the flawed claimant count measure to a full monthly survey of unemployment and employment. This option, preferred by many experts, would cost pounds 8m to pounds 10m a year for relatively little new information.

However, the ONS is expected to introduce a monthly measure of unemployment based on the current quarterly surveys, at an additional cost of around pounds 250,000.

Following consultations during the summer, it will also shape its presentation of the mass of information available on the jobs market to reflect better the concerns of the Government. In particular, it will give more prominence to the results of the detailed quarterly Labour Force Survey, merging them with the claimant count unemployment measure and other information such as earnings and vacancies.

As the ONS put it in its announcement of the consultation exercise, the point is to "allow the broad analysis of the labour market to stand out more clearly".

It is likely to introduce figures showing quarterly changes in unemployment for "a number of groups with differing degrees of attachment to the labour market", which could include groups of particular interest for Government policy, such as young people. There will also be an analysis of part-time and full-time employment, temporary versus permanent jobs, and economic inactivity by reason (such as full-time education).

The changes, due to be announced in the ONS's monthly publication, Labour Market Trends, will be welcomed by economists and other experts who rely on the figures. The ONS will describe the improvements as a matter of better presentation, but even so they will help to depoliticise one of the most important indicators of the health of the economy.

Although the new emphasis on the survey-based measure, which conforms to the internationally accepted standard, will "add" hundreds of thousands of people to the unemployment total, joblessness on both measures is currently falling rapidly.

The Government is still considering how to implement its manifesto pledge to create an independent official statistical service.